Power Supply Resevoir Size - Page 166 - diyAudio
Go Back   Home > Forums > Amplifiers > Power Supplies

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 11th December 2012, 05:58 AM   #1651
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Attached is the latest version of the PDF description of the spreadsheet and derivation of equations and numerical solution, etc. (The complete VBA code is available in the spreadsheet downloads.)

Some actual errors were corrected and some significant new material was added. Please delete any older versions, after downloading this version.

Cheers,

Tom
  Reply With Quote
Old 11th December 2012, 06:22 AM   #1652
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Quote:
Originally Posted by Krisfr View Post
I asked these questions months ago

If I wanted to make a 400 watt into 8 ohm/ 650watt into 4 ohm amp, what would the power supply specs be?
Should each rail have its own transformer and bridge rectifier and which is better

400 100uf caps
40 1,000uf caps
4 10,000uf caps
1 40,000uf cap

should each output transistor have a local cap and a bypass cap of about .1uf ?

Would anyone have an answer now that the spreadsheet is out?

Thanks for any positive answers...

I am trying to make a decision based upon total cost of a project
Krisfr,

The spreadsheet could give you a pretty-good idea of what would work, and how well, and would show any obvious trends or differences for the capacitor configurations. It could also show you what happens when a series resistance is inserted, for example (probably not a good idea- use one that gets switched out after the initial inrush, instead).

But your capacitor configuration questions would be better to investigate with a more-sophisticated type of tool, probably LT-Spice unless you have access to the really-expensive types of PCB-modeling software.

The reason is that with a very large number of parallel capacitors, the physical layout can easily make things like the inductance and resistance of the conductors that connect them become significant and there could be good and bad versions that are very different from each other, even for the same number and value of capacitors.

Your best bet is to take Abraxalito's advice and go back to Terry Given's posts about making a PCB full of caps for each rail. If you use his method, the calculations will be almost simple-enough to do in your head and you can then use a lumped model (a one-capacitor equivalent) with the spreadsheet, or LT-Spice, to explore further.

Regarding using decoupling and bypass caps for each transistor, I would tend to say yes (and either "definitely" or "why not?"), unless they are very close to the edge of a Terry Given capacitor bank.

Cheers,

Tom
  Reply With Quote
Old 11th December 2012, 09:22 AM   #1653
diyAudio Member
 
Join Date: Oct 2006
Default Well done for getting that far

Now do some listening tests - how much deviation from this calculated optimum can you go before it matters much?

Are these margins changed much for odd loads likes headphones, frequency specific speakers and weird driver technoilogy like electrostatics?

Compare a single channel with dual mono or multiple psus for single channel.

Will the spreadsheet work for powering the low power bit on an amp? often designs have the input stage and vas having some amount of additional capacitance that looks like a smoothing cap but is in the 200-500uf range.
  Reply With Quote
Old 11th December 2012, 11:09 AM   #1654
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
I think that every amplifier, whether low current opamp style or high current Power Amp style needs local decoupling.
That local decoupling performs better, the closer it is to the current changing device, usually the output device.
__________________
regards Andrew T.
  Reply With Quote
Old 11th December 2012, 01:40 PM   #1655
a.wayne is offline a.wayne  United States
diyAudio Member
 
a.wayne's Avatar
 
Join Date: May 2008
Location: Front Row Center
Agree, the magic trick is the value .......
  Reply With Quote
Old 12th December 2012, 03:11 AM   #1656
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Quote:
Originally Posted by a.wayne View Post
Agree, the magic trick is the value .......
From i = C x dv/dt (the standard ideal capacitor equation), you can use:

C ≥ (Δi x Δt) / Δv

where you pick or specify the Δv rail disturbance that can be tolerated, when a Δv is caused by a Δi being demanded/allowed to occur (by the pin or device being decoupled by the C), with the Δi starting and finishing within Δt seconds. (Note that longer Δt is the worse case for Δt, but the shortest required Δt can probably also be used, to find a value for a small parallel cap for high-frequency decoupling.)

Alternatively, you can look at Δv / Δi and say it's an impedance; the "target maximum impedance" that the pin or device must see not more than, at all frequencies up to some maximum. Then you can relate it to the capacitance's impedance, 1 / jωC, where ω = 2πf, i.e. angular frequency in radians per second, with f being frequency in cycles per second (Hertz).

Typically, for an audio amp, f would need to go up to somewhere between 80 kHz and 400 kHz. Edge times could also be used to help select f, with f = 1 / (π x trise).

Then, you must also find out what the maximum tolerable inductance is, which will mainly determine the allowable total length of the connections between the decoupling cap and the pins/device being decoupled. For that, you can start with the standard inductor equation, V = L x di/dt and use

L ≤ (Δv x Δt) / Δi

(Or maybe you can figure the L as part of the target maximum impedance. And ESR also gets accounted for.)

Any of the above information might be somewhat wrong, or at least incomplete, because it's been a long time since I worked on the decoupling capacitor calculations, and they weren't quite finished even back then. So please also see the links to some previous posts about it, at http://www.diyaudio.com/forums/solid...ir-size-3.html , or, from very early in this thread. Also recommended are Henry W. Ott's book, "Electromagnetic Compatibility Engineering" and some stuff on the web by Bruce Achambeault (spelling?).

You also usually need a high-frequency BYPASS cap, extremely close to the device pins, because most transistor amplifier toplogies have a hidden positive feedback loop through the power rail, which must be shorted for high frequencies to prevent instability, or marginal stability, which would cause oscillation or ringing at HF.


Last edited by gootee; 12th December 2012 at 03:16 AM.
  Reply With Quote
Old 12th December 2012, 01:25 PM   #1657
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Forgot to mention that the delta v on the rail is caused by the delta i trying to flow through the inductance of the rails. V = L di/dt so a faster-changing delta i will cause a larger delta v. The purposes of decoupling caps are to minimize the delta v rail disturbance and to deliver the delta i current accurately, when demanded, avoiding the delay time that the rail inductance would otherwise induce, which would result in distortion of the "shape" of the transient current's waveform, in terms of the amplitude vs time. The rail inductance lowers the high-end frequency response and the decoupling capacitance raises it.

Last edited by gootee; 12th December 2012 at 01:30 PM.
  Reply With Quote
Old 12th December 2012, 09:21 PM   #1658
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
I got the mouse-drag "ZOOM" functionality working, in the spreadsheet's plot, and thought it should be shared. Attached is beta version 0.99.

I am hoping that someone will let me know if it works in any version of Excel other than 2007.

I included an Excel 97-2003 xls file version, and the 2007 xlsm file, both in zipped folders.

I attached two screen images. In the first one, you can see the rectangle that I just dragged with my mouse, on the plot. The next screen image shows what happened when I released the mouse button.

Notice the three new buttons, just above the plot area. They are "AUTOFIT", "ZOOM", and "UNDO ZOOM".

The AUTOFIT button rescales the plot so that the axes will show every part of the plots. The ZOOM button makes it wait for you to drag a rectangle in the plot area, and then zooms-in so that rectangle fills the plot area, with correctly-scaled axes. The UNDO ZOOM button makes the plot go back to the previous zoomed state. It can handle multiple levels of zoom-undo.

I didn't write the code for those button actions, thankfully. But I did adapt it to use buttons instead of a menu bar (since CommandBars objects are obsolete in Excel 2007.) However, I have noticed that if you click and drag OUTSIDE of the plot area, which the code does NOT prevent, it will cause weird things to happen. So please avoid clicking outside of the plot area, just after pressing the ZOOM button.

Cheers,

Tom
Attached Images
File Type: jpg v0.99scrn_prezoom.jpg (419.5 KB, 302 views)
File Type: jpg v0.99scrn_postzoom.jpg (388.5 KB, 281 views)
Attached Files
File Type: zip Power Supply and Transformer Analysis v0.99_xlsm_2007.zip (903.8 KB, 20 views)
File Type: zip Power Supply and Transformer Analysis v0.99_xls_97-2003.zip (556.5 KB, 21 views)
  Reply With Quote
Old 13th December 2012, 11:03 AM   #1659
KSTR is offline KSTR  Germany
diyAudio Member
 
KSTR's Avatar
 
Join Date: Jul 2007
Location: Central Berlin, Germany
Tom, seems to run fine on 2003, except for the zooming stuff (but maybe operator error).
  Reply With Quote
Old 14th December 2012, 12:08 AM   #1660
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Quote:
Originally Posted by KSTR View Post
Tom, seems to run fine on 2003, except for the zooming stuff (but maybe operator error).
Thanks very much for the feedback, Klaus!

Would I be correct in assuming the the four scroll-bars do work, for "manual zooming"? All they do is change the plot axes' maximum and minimum values.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Valve power supply - How to size transformer? SanderW Power Supplies 25 4th January 2013 04:12 PM
How do you calculate choke size in a power supply? Original Burnedfingers Tubes / Valves 25 5th January 2012 12:23 AM
power supply bypass cap size BigE Power Supplies 11 5th July 2011 02:59 PM
Power Supply Case Size diymixer Power Supplies 1 10th October 2010 05:47 AM
What size power supply should I get for repair work? spooney Car Audio 3 6th December 2007 11:50 PM


New To Site? Need Help?

All times are GMT. The time now is 08:29 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2